A predictive resampler scheduler algorithm is a process for modifying the size of an audio buffer. In some situations, an audio encoder may be operating at a different clock speed from a coupled audio decoder. For example, in an audio provider/consumer environment, every time the producer's clock ticks, a new audio frame has become available. Every time the consumer's clock ticks, it is hungry for one audio frame. In conventional systems, if the producer's clock ticks faster than the consumer's, too much data may accumulate and overrun the consumer's buffers that may result in skips in the audio output. If the producer's clock ticks slower than the consumer's, the consumer will be starved for data and may experience gaps in the audio output.